I've stirred up a bit of controversy in the past by asking whether there's really such a thing as negative links (as I said before, the answer's complicated), and more specifically if they play a part in Penguin (my guess is no). But the reason I took these stances wasn't just to start a debate about the algorithm. It was primarily because of my growing contempt for a branch of SEO called "link removal," "link detox," or what have you.
As you might sense from my tone, I don't like link removal. But should you ever bother with it? Well, I'm going to plant my flag firmly in the ground and proclaim that, once again, it's complicated. The real world always is.
Let's Start With the Pure SEO Perspective
There are basically just two situations where you should find yourself thinking about link removal from a pure SEO standpoint:
- You have been manually penalized by Google for unnatural links and, as a result, you've received a message about it in Google Webmaster Tools.
- An algorithmic update like Penguin has reduced your rankings.
Let's tackle those one by one.
When it comes to links, Google has three manual penalties you can end up with:
- A site-wide penalty for unnatural inbound links
- A partial penalty for unnatural inbound links
- A penalty for unnatural outbound links
The third one is a relatively easy fix since you can remove the links yourself and get the penalty reversed. The first two are quite a bit more difficult, and require some deliberation.
There is not a shadow of a doubt in my mind that you can improve your rankings simply by getting a site-wide penalty for unnatural inbound links revoked. Of course, you shouldn't expect your rankings to return to their previous levels. Here's an example of what you can expect:
But partial link penalties are much vaguer. According to Google's own guidelines:
If you don't control the links pointing to your site, no action is required on your part. From Google's perspective, the links already won't count in ranking. However, if possible, you may wish to remove any artificial links to your site and, if you're able to get the artificial links removed, submit a reconsideration request. If we determine that the links to your site are no longer in violation of our guidelines, we'll revoke the manual action.
On top of that, a survey over at Search Engine Roundtable reveals that only 12 percent of webmasters see their rankings improve within days of getting a (non-specific) penalty revoked:
Things get more complicated. I previously believed that partial link penalties simply removed links, rather than counting them against you, but further research suggests that this may not always be true. While Google's own guidelines say that the links already don't count in ranking, Matt Cutts originally said that they might prevent you from ranking for specific phrases:
If you do a search for [widgetbait matt cutts] you'll find tons of stories where I discourage people from putting keyword-rich anchortext into their widgets; see http://www.stonetemple.com/articles/interview-matt-cutts-061608.shtml for example. So this message is a way to tell you that not only are those links in your widget not working, they're probably keeping that page from ranking for the phrases that you're using.
Go figure. It seems that the partial link penalty is actually a label for several different penalties, some of which might count against you, and some of which might only ignore the links, depending on the situation.
Nevertheless, rebounds from partial link penalties, if they occur at all, aren't especially dramatic. Take this example from Koozai:
This is fairly typical among success stories. Since we expect success stories from SEO firms to represent the best scenarios out there, it's not a good idea to hope for much after a partial link penalty is revoked. Remember, only 12 percent see any improvement at all within days. In my opinion, anything that happens after that has more to do with your additional SEO efforts, or random chance, than the removal of the penalty.
Penguin and other Algorithmic Demotions
If you followed the link at the beginning of this post, you already know that I have my doubts about negative links within Google's algorithm.
When Matt Cutts said that the original Penguin only looked at the homepage of a site, he was almost certainly talking about the homepages of sites with unnatural outbound links, otherwise he was directly contradicting data demonstrating that Penguin did indeed look at other pages on your site.
Matt Cutts has also said that Google "typically" ignores outbound links from link sellers, as opposed to penalizing the sites receiving links from those pages. Warnings from Matt Cutts are also typically aimed at sellers, not buyers.
Furthermore, when Specer Haws was hit by Penguin 2.1, it took his traffic levels down to about 400 visitors per day, the same level of traffic he saw immediately before the negative SEO attack began. Several other examples corroborate this.
As I've said before, it doesn't make sense from the perspective of a Google engineer to algorithmically penalize sites for suspicious inbound link profiles. The risks of a false positive are too high, and it's contrary to the way Google crawls the web. It makes more sense for Google to engineer an algorithm that identifies suspected link sellers, and simply ignores outbound links from them. This maximizes the impact on manipulative sites while minimizing the impact on legitimate sites. The data I've seen is better supported by this theory, at least in my opinion.
Nevertheless, I have to admit that examples like this from Marie Haynes are a possible challenge to this theory:
While I personally believe that these are just changes rippling through the link graph, cutting more link value from Marie's competitors than from her client's site, I'd be unnecessarily stubborn if I didn't admit that the link removal might have helped improve rankings. It's worth noting that this is the most stark recovery from Penguin I've ever seen. Most are very modest, nearly indistinguishable from trends and random fluctuation. This is likely the very best case in Marie's database of who knows how many clients.
The real question is whether or not it's worth the effort to remove links when you won't see results until the next Penguin update, you have no idea when that will occur, and you can't be sure if any recovery you see actually has anything to do with the link removal.
The Politics of Link Removal
Julie Joyce has pointed out that you can only expect a 10 percent success rate with outreach. Some of her clients have had months where they sent out 1,000 requests and successfully removed zero of their links. On top of that, some webmasters are asking for between $20 and $500 for a link removal.
Many sites that have been either penalized or algorithmically demoted have thousands or even tens of thousands of spammy links to deal with.
It's worth asking whether it's a better use of your time to send thousands of emails asking for links to be removed, or to send thousands of emails trying to build relationships and the links that come with them.
As far as my particular tastes go, there's only one situation where it makes sense to go through that kind of an effort: if your site has received a site-wide manual penalty.
Even then, I would only consider it worthwhile if your site has a non-search engine audience, or if it has enough legitimate links to support good rankings after the links have been removed. Otherwise, I would say start from scratch.
As for partial link penalties and Penguin, I'll admit it's possible that link removal may improve rankings, but I suspect your efforts are best spent elsewhere. I would suggest investing in building links from highly influential places on the web, strengthening your existing audience, and finding other sources of referral traffic. If you discover that these efforts aren't having the kind of impact you expect with the search engines, perhaps then you can revisit the topic of link removal, and test the theory that your rankings are truly being held back.
Just keep in mind that even if you get the penalty removed, 88 percent of you won't see an improvement in rankings unless you start building links. I'd start with building links.
Maybe that's just me.